摘要
动态过程风险评估需要综合考虑时间、过程变量、元件的变化及其耦合，要求用动态的方法来模拟这种耦合并估计动态过程的可靠性．介绍了描述动态随机过程的一般性理论——概率动力学．对求解概率动力学的两类主要方法：ＭＣ仿真和ＤＹＬＡＭ的优缺点作了讨论．提出了一种综合这两种方法优点的基于离散事件树的ＭＣ仿真方法．最后以一个算例说明其良好的效率和精度．
Through a typical example, the disabilities of conventional ET/FT(event tree/fault tree) methodology are explained. Probabilistic dynamics, which is a general modeling theory for dynamic stochastic processes, is introduced. CET(continuous event tree) based MC (Monte Carlo) simulation and DET(discrete event tree) based DYLAM(dynamic logic analysis methodology) are two of the most general methods for solving probabilistic dynamics questions. A new method, DET based simulation approach, is put forward, which synthesizes the advantages of the two methods. A calculation example is given.
出处
《上海交通大学学报》
EI
CAS
CSCD
北大核心
1998年第11期7-20,25,共15页
Journal of Shanghai Jiaotong University
关键词
概率动力学
风险评估
动态过程
随机过程
probabilistic dynamics
continuous event tree
discrete event tree
risk assessment
